Details on the value of data linkage in research and policy formulation can be found in a report from the National Research Council (1988) and other sources (Kelman and Smith, 2000). A corollary issue is the need to determine the accuracy and completeness of the records being linked.

An unexplored area in determining elder mistreatment occurrence is the application of biomarkers. A biomarker in this case is any physical, physiological, or biochemical measure that could assist in identifying victims of elder mistreatment and could most easily be acquired in field surveys via blood or urine specimens. Even if biomarker associations with elder mistreatment are proven, these are much more likely to indicate increased risk and would not lead to definitive elder mistreatment designation. Some biomarker applications may relate to undernutrition, such as blood cholesterol, albumen, or micronutrient levels. Others may relate to chronic psychological or physical stress, but as blood or urinary catecholamine or cortisol levels or markers of chronic immune dysfunction. Chronic blunt trauma may increase blood or urinary myoglobin or other muscle protein degradation products. Additional forensic techniques, both antemortem and postmortem, may be useful detecting elder mistreatment cases. It is not outside the realm of possibility that genetic markers may be candidates for elder mistreatment research, to the extent that they perhaps reflect particular behaviors, diseases, or responses to stress and trauma. As one example, somatic mutation rates in the genome have been proposed as an indicator of cumulative environmental exposures (Albertini, 1998). The application of biomarkers to elder mistreatment assessment could be an area for possible future research. A recent volume addresses many aspects of applying biomarker acquisition to population surveys (National Research Council, 2000b).

1. Population-based surveys of elder mistreatment occurrence are feasible and should be given a high priority by funding agencies. Preparatory funding should be provided to develop and test measures for identifying elder mistreatment.

There is inadequate information on elder mistreatment occurrence among both community-dwelling and institutionalized elders. However, before embarking on such surveys, the aims and rationale for them should be clearly delineated, and the strengths and weaknesses of the survey methodology fully understood. Different methods and approaches may be required for various types of mistreatment, and multiple modes of case ascer-